Agreed. Also what about generating X groups of randomised numbers of hashtags?

For example, if we grouped hashtag collections by:

And we want Flick to generate (example 100) groups of between [5] and [9] hashtags ... we can select which Collections to pull from and specific what percentage to use ...

Example: 65% [Low Comp], 25% [Medium Comp] and 10% [High Comp]

That way we'll have completely randomised groups of relevant hashtags to pull from but using solid "staircasing" strategy.

without that, mobile app and groth plan is useless for me, thanks